In partnership with

Speak fuller prompts. Get better answers.

Stop losing nuance when you type prompts. Wispr Flow captures your spoken reasoning, removes filler, and formats it into a clear prompt that keeps examples, constraints, and tone intact. Drop that prompt into your AI tool and get fewer follow-up prompts and cleaner results. Works across your apps on Mac, Windows, and iPhone. Try Wispr Flow for AI to upgrade your inputs and save time.

If you’d like to pay it forward today, you can share this content and/or buy me a coffee, thank you for supporting this work ☕

Hi fellow humans,

Every account, every click, and every password is a doorway, and the wrong one left open can upend real lives and expose people’s true reality and beliefs, the good and the evil. Now AI is also involved.
I’m reading tons now on cybersecurity and its human factors, and I’m amazed how naive humans can be, even if they are very careful and cautious; in fact, it is not really naivety; it is humanity and its vulnerabilities.
Today’s story took place two months ago, but the aftermath is still unfolding, one story at a time. “White leaks” is about online extremism, a 44-minute talk, and a researcher in costume and artificial intelligence. But at its core, it all comes back to a difficult question: What is the future of Cybersecurity? Is AI taking this ongoing cyber war to the next level?  By the end, you may find yourself weighing that tension too.

We’ll walk through what happened together, checking the primary sources. And you can decide what to think.
A sneak peek > During one of the world’s biggest security conferences, a researcher shared how she spent months using AI chatbots to track extremist platforms. Then…Read what happened. And in the aftermath, UK and German politicians were exposed!

Wafaa Albadry

HUMAN THINKING
Return to the top of the page.
What Happened After the Pink Power Ranger Deleted the Far‐Right Dating Apps

PART ONE

The Platforms

At the end of December 2025, in Hamburg, Germany. The 39th Chaos Communication Congress, one of the world's largest annual gatherings of security researchers, technologists, and digital rights advocates, is in its third day. A speaker steps onto the stage. She’s dressed head-to-toe in a pink suit and helmet, just like the Pink Ranger from Power Rangers. The audience has no idea what’s coming. The panel is named "The Heartbreak Machine: Nazis in the Echo Chamber”

Over the next 44 minutes, she will reveal that she has spent months on three interconnected far-right dating and networking platforms. She will explain how she deployed AI-powered chatbots to interact with users. She will show the data she collected.

The researcher uses the name "Martha Root," taken from Martha Louise Root (1872–1939), an American Journalist and Baháʼí Faith educator who travelled the world promoting peace and unity across religious and national lines.

According to the official CCC event listing, Root spent months researching three connected websites, all run by the same person. The administrator of white supremacist platforms WhiteDate, WhiteChild, and WhiteDeal is allegedly identified as a woman based in Kiel, Germany, and has previously worked as a pianist.

The three platforms:

WhiteDate described itself as a "dating platform for white people with traditional European values." Journalist Eva Hoffmann, who presented with Root, identified it as a white supremacist dating site. The site had between 6,500 and 8,000 registered users, mostly from the United States, Germany, the United Kingdom, Canada, and France.

WhiteChild focused on what it called "family and lineage ideals,"connecting users seeking racially segregated reproduction services.

WhiteDeal was a networking and services marketplace for white supremacists.

Root set out to document how these sites worked.

━━━━━━━━━━━━━━━━━━━━

PART TWO

The Chatbots

Root’s approach combined two things: the platforms’ security weaknesses and AI's conversational abilities.

First, security. During her talk, Root showed that WhiteDate’s user database could be accessed via a simple URL path, essentially a download button out in the open. She called the platforms’ security practices minimal.

But Root went further. She deployed AI chatbots directly into the community.

According to the CCC event description, Root "introduced 'realistic' AI chatbots into these spaces. The bots were convincing enough to bypass verification checks and even earn 'white verified' status, leading some users to develop emotional attachments."

The bots ran on Ollama with locally hosted language models. Root created personas that were designed to blend in with the platforms' user base, creating profiles that showed interest in what the platforms called "traditionalist values."
One of Root’s accounts was suspended on suspicion of using a bot. She was asked to “white verify,” which required her to open the camera. She used a live deepfake for that.

According to Root's presentation, one of her chatbot accounts, operating under the username "lilmisethnostate", was eventually invited to an in-person meetup in northern Germany by a user called "Anglo-Saxon."

Screenshot of Martha Root's presentation from the 39th Chaos Communication Congress (39C3).

Root worked alongside journalists Eva Hoffmann and Christian Fuchs from the German publication Die Zeit. They had published an investigative article about the platforms in October 2025, and both appeared onstage with Root at the CCC presentation.

━━━━━━━━━━━━━━━━━━━━

PART THREE

The Presentation

For 44 minutes, Root, still in her pink costume, explained everything: how she found the platforms, built and used the chatbots, what she learned about how the sites operated, and who was behind it.

Then came the last few minutes.

Root opened a terminal window on her laptop and ran a Python script. The presentation said the script was called "lol.py."

The audience watched as text scrolled across the screen:

Martha Root deleting the platforms, screenshot from the 39th Chaos Communication Congress (39C3).

The presentation also said Root changed the administrator’s account credentials and took down the platform’s social media. The administrator’s X account was later restored. Recently, the admin has stated that the Dating website will be back, but they “are still testing and debugging.”

━━━━━━━━━━━━━━━━━━━━

PART FOUR

The Data

The platforms went offline, but Root had already collected data.

Her documentation says she extracted about 100 gigabytes of data, including over 8,000 user profiles, photos (some with GPS coordinates), biographical details, and stated beliefs.

The demographic breakdown, per Root's analysis: 86% male, 14% female.

Root published some of this data on okstupid.lol, a website she created with an interactive map of user locations (from photo metadata), as well as redacted images and pseudonymized profiles.

The complete dataset was sent to Ddosecrets, an organisation that archives leaked datasets. They titled the release "WhiteLeaks" and restricted access to verified journalists and researchers.

Screenshot from Okstupid.lol showing leaked EXIF data revealing users’ real addresses, captioned: “3,831 users with geotagged photos. Click the potatoes to see their profiles.”

━━━━━━━━━━━━━━━━━━━━

PART FIVE

The Aftermath

The administrator of the three platforms, who operates publicly under the pseudonym 'Liv Heide', publicly condemned the action as "cyberterrorism" and vowed retaliation, though no formal charges have been announced by police. The administrator responded on X:

"They publicly delete all my websites while the audience rejoices. This is cyberterrorism. No wonder some of them hide their faces. But we will find them, and trust me, there will be repercussions."

German authorities have not issued public statements regarding the incident.  German media and legal commentators have noted that while the hack was celebrated at the Chaos Communication Congress (39C3), deleting servers and leaking data is "legally problematic" and technically constitutes a criminal offence under German law. This has led to speculation on platforms like Reddit about whether German authorities will choose to investigate Root or focus on the extremist activities revealed by the leak.

According to the CCC event page, the platform had around 8,000 members across 88 countries. The published sample on okstupid.lol showed roughly 400 users each from Germany, the UK, and France. With Germany at 10% and, the UK at 7%, and the majority in the US.

Political Fallout in the UK

The leak crossed into politics quickly. The Observer reported that around 450 of the 8,000 profiles were from the UK, including, according to the newspaper, former members of the British National Party and activists associated with Patriotic Alternative.

Two local councillors have been publicly named.

Lillith Osborn, a Conservative representing Glastonbury Town Council since 2022, was suspended by her party after her name appeared in the database. Her profile reportedly described her as "pro-white." Osborn told The Observer she had visited the site "once or twice" while researching modern paganism, and that she has mixed-race children. The Conservative Party confirmed she is suspended pending investigation. Liberal Democrat MP Sarah Dyke called for her resignation, describing WhiteDate's associated views as "utterly vile."

Gavin Beales, a Reform UK councillor in North Northamptonshire, was also named. He stated he "may have inadvertently been linked" to the site years ago and "cannot clearly recall the details." Three Labour MPs called for his suspension. Reform UK said it found "no case to answer."

Political Fallout in Germany

The journalistic investigation began before the CCC presentation. In October 2025, Die Zeit journalists Christian Fuchs and Eva Hoffmann published an investigation into WhiteDate and its operator, a woman from Schleswig-Holstein.

The North Rhine-Westphalia Office for the Protection of the Constitution (Verfassungsschutz) had already reported briefly on the platform in 2019, classifying it as "extremism in digital space."
The data touched nearly every corner of Germany's extremist landscape. According to journalist Christian Fuchs, speaking at the CCC presentation, users included five local AfD politicians, three members of the BSW, at least two members of the Identitäre Bewegung, a woman living on a farm linked to the Anastasia movement, and one AfD politician in a state parliament.

Context for international readers:

  • AfD : Far-right party; identified as extremist by BfV (May 2025), label suspended by court (Feb 2026)

  • Identitäre Bewegung : Far-right movement; confirmed extremist by BfV since 2019; promotes the "Great Replacement" theory

  • Anastasia Movement : Esoteric sect; monitored as suspected extremist for völkisch nationalism and antisemitism

  • BSW: Left-wing populist party; socialist economics + cultural conservatism + restrictive migration policy. The BSW (Sahra Wagenknecht Alliance), however, is a left-wing populist party; its members' presence on a white supremacist dating site is unexpected and unexplained.

A TAZ report identified an AfD member of the Hamburg state parliament (Bürgerschaft) in the leaked database. His username was "Waldgänger" - allegedly a reference to Ernst Jünger's essay "Der Waldgang," Jünger being an icon of Germany's New Right.

The AfD politician has not publicly commented. The AfD has not issued a statement. As of publication, no disciplinary measures have been announced. He remains in office.

However, though the story broke in Germany—German journalists, a German hacker, a German-run platforms—the political consequences have so far played out elsewhere. In the UK, one councillor was suspended within days. In Germany, a state parliamentarian remains in office.

The full dataset remains restricted to journalists and researchers via DDoSecrets. More names, potentially from other countries, may emerge.

The Bigger Picture: This Time, AI Did the Flirting

This isn't the first time extremist platforms have been exposed. In 2021, hackers claiming to be affiliated with Anonymous announced they had infiltrated Epik, a web host that serves Gab, Parler, and similar sites. The leak reportedly exceeded 180 gigabytes. Extremism researcher Megan Squire called it "the biggest domain-style leak I've seen." Weeks later, membership records attributed to the Oath Keepers surfaced via DDoSecrets, over 38,000 names, including individuals believed to be police officers, military members, and elected officials. Some faced internal reviews after being named.

Doxxing cuts across political lines. In 2017, The Intercept reported on neo-Nazi-affiliated Discord servers where users compiled personal information on left-wing activists and journalists, not to inform, but to intimidate.

White supremacist movements do not stay online. Their members teach, police, legislate, and serve. We need to draw a clear line between exposure for accountability and doxxing for harassment. Exposing public figures who use their power to advance hateful ideologies is an accountability leak. By contrast, harassment doxx dumps aim to intimidate, silence, or endanger private individuals, often by publishing information with no public interest. This exposure is a form of accountability, but WhiteDate adds something new.

Previous leaks exploited technical failures and weak passwords, exposing databases. This one exploited that, but it also focused on the psychology of a conversation through an AI Chatbot. Root used large language models running locally via Ollama. The bots generated contextually appropriate responses, mirrored beliefs, and maintained personas over months. In social engineering terms: automated rapport-building at scale. One chatbot, "lilmisethnostate," was invited to a real-world meetup.

The CCC framed Root's project as showing "how AI personas and investigative thinking can illuminate extremist networks." This appears to be the first documented case in which AI was used not to scrape an extremist platform but to socially infiltrate it.

The next generation of researchers and journalists will need to understand how AI builds trust, how verification fails, and how hate operates in plain sight. AI literacy and in-depth knowledge are no longer optional.

Three platforms that promoted white supremacist ideology are offline. Over 8,000 profiles are with journalists and researchers. The researcher in the pink costume has left the stage.

The question is who learns these methods next, and for what purpose.

(A note on legality: The legal status of Root’s actions under German, European, or international law has not been publicly decided. Laws about computer access, data collection, and website disruption differ by country. We do not judge legality; that’s for courts and legal experts.

As of publication, all three websites remain offline.

• • •
That’s this Month’s story: finding the human signal in the noise.

See you next Month 💙

Wafaa

━━━━━━━━━━━━━━━━━━━━

Important Notice

AI.Human.Story reports on artificial intelligence, technology, and their impact on society. Our coverage documents real events, and emerging debates.

We do not endorse or encourage unlawful conduct of any kind, nor do we support extremist or hate-based ideologies. Reporting on actions, individuals, or viewpoints does not imply approval.

Content is provided for informational purposes only and does not constitute legal advice.

We believe in human agency and informed judgment. Readers are responsible for their own decisions and, where appropriate, should consult qualified professional counsel.

AI.Human.Story assumes no liability for actions taken based on its published material.

AI Use Disclosure
All editorial decisions and texts are human written. We use tools such as Grammarly and Claude by Anthropic for limited editorial support and occasional visual assistance. These tools assist our workflow but do not replace human judgment, accountability, or authorship. We stand for human craft and responsible innovation.

HUMAN THINKING

Dispatches on AI, technology, psychology, and what it means to be human.

Have a story tip? Reply to this email.

Get involved

This Newsletter aims to be created by those who see the human signal in AI noise. We plan to publish reader submissions. 

We want your opinion and if you’ll share or repost your writing, academic paper, reflections, collaboration, or contribution.

Disclaimer: All content reflects the views of the authors and not those of the publisher, which maintains full editorial independence and assumes no liability for author statements.

1 

Keep Reading